| --- |
| title: Email Classification For Support Team |
| emoji: ⚡ |
| colorFrom: green |
| colorTo: yellow |
| sdk: docker |
| pinned: false |
| license: mit |
| --- |
| # Email Classification API for Support Teams |
|
|
| ## Introduction |
| This API classifies support emails into four categories (Incident, Request, Problem, Change) using a fine-tuned DeBERTa model. The service automatically masks personally identifiable information (PII) before processing, ensuring privacy compliance. |
|
|
| ## Hardware Requirements |
| - **For Development/Testing:** |
| - CPU: 2+ cores |
| - RAM: 4GB minimum (8GB recommended) |
| - Storage: 1GB free space |
|
|
| - **For Production Deployment:** |
| - CPU: 4+ cores |
| - RAM: 8GB minimum (16GB recommended for higher throughput) |
| - GPU: Optional but beneficial for faster inference |
|
|
| ## Software Requirements |
| - Python 3.8+ |
| - Flask |
| - PyTorch |
| - Transformers library |
| - PII masking utilities |
| - Internet connection for initial model download |
|
|
| ## How to Run |
|
|
| ### Local Development |
| ```bash |
| # Clone repository |
| git clone <repository-url> |
| cd email-classification-api |
| |
| # Install dependencies |
| pip install -r requirements.txt |
| |
| # Run the application |
| python app.py |
| ``` |
|
|
| ### Hugging Face Spaces Deployment |
| 1. The API is already deployed at: https://huggingface.co/spaces/Nikpatil/Email_Classification_For_Support_Team |
| 2. No additional setup required to use the deployed version |
|
|
| ## How to Access |
|
|
| ### Using the API Endpoint |
| Send POST requests to: `https://nikpatil-email-classification-for-support-team.hf.space/classify` |
|
|
|
|
|
|
